Scribd
Upload
165 views20 pages

Econometrics

This document summarizes key concepts from Lecture 5 of an econometrics course, including: 1) Log-linear (log-log) regression models that allow for nonlinear relationships by taking logarithms of variables. 2) Using dummy variables to control for qualitative factors like gender in regressions and test for statistically significant differences. 3) Partial adjustment models that incorporate lagged dependent variables to capture dynamic adjustment effects over time. 4) An example study of cigarette consumption in Greece that illustrates applying these techniques.

Uploaded by

Mia Caldwell
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
165 views20 pages

Econometrics

This document summarizes key concepts from Lecture 5 of an econometrics course, including: 1) Log-linear (log-log) regression models that allow for nonlinear relationships by taking logarithms of variables. 2) Using dummy variables to control for qualitative factors like gender in regressions and test for statistically significant differences. 3) Partial adjustment models that incorporate lagged dependent variables to capture dynamic adjustment effects over time. 4) An example study of cigarette consumption in Greece that illustrates applying these techniques.

Uploaded by

Mia Caldwell
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 20

Introduction to Econometrics

Lecture 5
Extensions to the multiple regression model

INEMET [U13783]

Guy Judge March 2007

Lecture plan
logarithmic transformations - log-linear

(constant elasticity) models


dummy variables for qualitative factors
simple dynamic models with lagged variables the partial adjustment mechanism
an application to illustrate the above - A study
of cigarette consumption in Greece by Vasilios
Stavrinos (Applied Economics, 1987 pp 323329)
INEMET [U13783]

Guy Judge March 2007

Log-linear regression models (1)


In many cases relationships between
economic variables may be non-linear.
However we can distinguish between
functional forms that are intrinsically nonlinear (and will need to be estimated by
some kind of iterative non-linear least
squares method) and those that can be
transformed into an equation to which we
can apply ordinary least squares
techniques.
INEMET [U13783]

Guy Judge March 2007

Log-linear regression models (2)


Of those non-linear equations that can be
transformed, the best known is the
multiplicative power function form
(sometimes called the Cobb-Douglas
functional form), which is transformed into
a linear format by taking logarithms.

INEMET [U13783]

Guy Judge March 2007

Log-linear regression models (3)


Production functions
For example, suppose we have cross-section
data on firms in a particular industry with
observations both on the output (Q) of each firm
and on the inputs of labour (L) and capital (K).
Consider the following functional form

INEMET [U13783]

Guy Judge March 2007

Log-linear regression models (4)

INEMET [U13783]

Guy Judge March 2007

Log-linear regression models (5)

INEMET [U13783]

Guy Judge March 2007

Log-linear regression models (6)

The parameters and can be estimated directly from a


regression of the variable lnQ on lnL and lnK

INEMET [U13783]

Guy Judge March 2007

Log-linear regression models (7)

INEMET [U13783]

Guy Judge March 2007

Log-linear regression models (8)

INEMET [U13783]

Guy Judge March 2007

Dummy variables (1)


Dummy variables (sometimes called dichotomous variables) are variables that are
created to allow for qualitative effects in a regression model.
A dummy variable will take the value 1 or 0 according to whether or not the
condition is present or absent for a particular observation.
For example suppose we are investigating the relationship between the wage (Y)
and the number of years of experience (X) of workers in a particular industry.
Our initial model is
Y=a+bX+u
However we are concerned that the wages of female workers may be below that of
male workers with similar experience. To test for this we can introduce a dummy
variable to distinguish between the observations for male and female workers in
the regression.
INEMET [U13783]

Guy Judge March 2007

Dummy variables (2)


Define D = 1 for male workers and 0 for female workers.
The overall equation becomes
Y = a + b X + cD + u
where c will measure the differential between male and female
workers, having taken account of differences in experience. We can
run a normal multiple regression with X and D as explanatory
variables. Assuming that c is positive it means that the regression
line for male workers lies above that for female workers - c
measures the extent of the upward shift. We can use its t value to
test whether these differences are statistically significant.
INEMET [U13783]

Guy Judge March 2007

Dummy variables (3)


Ramu Ramanathan (1998) includes a data set compiled by Susan Wong relating to
49 professionals in an industry (23 are for females and 26 for males).
The results show a large and significant difference in wages (which range between
981 and 3833 with a mean of 1820).

INEMET [U13783]

Guy Judge March 2007

Dummy variables (4) Testing for differences in intercept.

Yi = 1 + 2 Xi+ 3 Di + ui
Y

wage
rate

For men: Di= 1.

Yi = (1+ 3) + 2 Xi + ut

Men
Women
For women: Di = 0.

Yi = 1 + 2 Xi + ui
1 + 3
1
0
INEMET [U13783]

years of experience

Guy Judge March 2007

Interactive dummies: Testing for differences in intercept and slope

Yi = 1 + 2 Xi + 3 Di + 4 Di Xi + ui
Y
wage
rate

Yi = (1 + 3) + (2 + 4) Xi + ui
Men

+ 4

Women
Yi = 1 + 2 Xi + ui

1
1 + 3
0
INEMET [U13783]

years of experience
Guy Judge March 2007

Dummy variables and time series data


With time series data we can have
impulse dummies just affecting a particular period
step dummies affect remains on for a number of periods
We might also have seasonal dummies
e.g. lnQt = b0 + b1 lnYt + b2lnPt + d1D1t + d2D2t + d3 D3t + ut
D1 = 1 for quarter 1 observations, 0 otherwise
D2 = 1 for quarter 2 observations, 0 otherwise
D3 = 1 for quarter 3 observations, 0 otherwise

Beware of the dummy variable trap


INEMET [U13783]

Guy Judge March 2007

Partial adjustment mechanisms (1)

INEMET [U13783]

Guy Judge March 2007

Partial adjustment mechanisms (2)

INEMET [U13783]

Guy Judge March 2007

Illustration: cigarette consumption in Greece


(see Stavrinos, Applied Economics, 1987 19, pp323-329)

INEMET [U13783]

Guy Judge March 2007

Stavrinos results

INEMET [U13783]

Guy Judge March 2007

You might also like